The gap the research uncovered

Over the past two decades, resuscitation researchers have repeatedly studied a deceptively simple question: when actual humans do CPR, how close does their technique come to the target standards? The targets are well-defined — current guidelines specify compression depth of 5–6 cm for adults, a rate of 100–120 compressions per minute, full chest recoil between compressions, and minimal interruptions.

The consistent finding across the literature: without real-time feedback, CPR providers — including trained healthcare professionals — frequently miss these targets. Compression depth is the most common miss, with providers tending toward too-shallow compressions. Rate misses tend toward too-fast (which compromises recoil and effectiveness). Hand position drifts over the course of long resuscitations. Compression-only ratios get worse as providers fatigue.

This isn’t a criticism of providers — it’s a recognition that CPR is a physically demanding motor skill performed under stress, and humans aren’t reliable measurement devices for their own technique. Without external feedback, even people who’ve done the training don’t know how their performance compares to the target in real time.

What current guidelines say about feedback

The American Heart Association and the International Liaison Committee on Resuscitation (ILCOR) — whose guidelines the Heart and Stroke Foundation of Canada follows — explicitly recommend the use of real-time audiovisual feedback during CPR training to improve compression quality. The guidelines reflect the cumulative evidence that feedback during training improves skill performance, and that feedback during actual resuscitation events can help maintain high-quality CPR through the duration of the event.

This recommendation has practical consequences. Major training manikins from Laerdal (the QCPR product line) and Zoll (Real CPR Help) include integrated feedback systems that show compression depth, rate, and recoil in real time on a screen or LED display. These manikins are standard equipment at BLS, ACLS, and PALS courses across Canada and the US.

Life Safe uses feedback-enabled manikins in our paid first aid and CPR courses — students see real-time compression depth and rate during practice, which is part of why hands-on instructor-led training is more effective than video learning alone.

How fast does CPR skill decay?

The other consistent finding in the research: CPR skills degrade meaningfully within months of training, even among providers who initially demonstrated good technique at certification.

Multiple studies have measured skill performance at various intervals post-training. The general pattern:

• Immediately post-training: most providers meet target depth, rate, and recoil
• 3 months post-training: meaningful decay begins; some providers fall outside targets
• 6 months post-training: substantial decay; many providers fall outside targets without realizing it
• 1 year post-training: significant population of providers performing well below target

This is part of the rationale behind annual BLS recertification for healthcare workers (vs the 3-year public CPR cycle) — clinical environments need higher confidence that providers’ skills haven’t drifted too far from baseline. It’s also part of the rationale behind in-between practice tools like feedback manikins, simulation training, and consumer apps like CPR Coach.

What feedback specifically improves

The research has been refined enough over the years to identify which aspects of compression performance specifically benefit from real-time feedback:

Why this matters for outcomes

Improved compression quality isn’t just a metric — it correlates with patient outcomes in cardiac arrest. The chain from feedback during training to better outcomes during actual arrests runs through several intermediate steps:

1. Feedback during training produces providers who can hit targets reliably under controlled conditions
2. Higher initial skill produces less rapid decay (you have to fall further to fall below the threshold)
3. Feedback during actual resuscitation events helps maintain quality during the high-stress real environment
4. Higher-quality CPR — closer to target depth, rate, and recoil — correlates with better return of spontaneous circulation rates and improved patient survival

Each link in this chain has accumulated evidence in the resuscitation literature, though the strength of evidence varies by link. The strongest evidence is for the training-quality link; the survival-outcome link is more difficult to establish because of all the confounding factors in real-world cardiac arrest.

The implication for consumer CPR practice

Most of the research above comes from healthcare and clinical training contexts. Healthcare providers use feedback manikins during structured courses; clinical resuscitation teams have access to feedback technology during actual cardiac arrest events. What about everyone else?

The general public — parents, community members, workplace first aiders, students — typically receives CPR training without real-time feedback during retention practice. They learn the technique once, get a certificate, and rely on memory and occasional refresher exposure until they either recertify or encounter a real emergency. Skill decay between courses is uncorrected.

This is the specific gap CPR Coach is designed to address. By bringing the feedback mechanism — depth, rate, recoil, positioning — into a consumer-accessible format that runs on existing phones, CPR Coach extends the benefits of feedback-enabled practice beyond formal training into between-course retention practice.

How CPR Coach implements feedback differently from clinical manikins

Clinical feedback manikins use built-in pressure sensors to measure force applied to a chest cavity. CPR Coach uses computer vision tracking hand position over time to estimate compression depth.

The two approaches measure different things, but both target the same outcome — letting the user know in real time whether their compressions are at the target standards. For clinical training where precision matters and budget is available, pressure-sensor manikins are the gold standard. For at-home consumer practice between courses, computer vision running on a phone is the accessible alternative that delivers the same category of feedback at zero cost.

See our technical explainer for how the computer vision side specifically works.

What we still don’t know

Honest framing: the research strongly supports feedback as an effective training tool. The research on consumer-grade feedback apps (where measurement is camera-based rather than pressure-based) is much newer and the evidence base is smaller. We can confidently say:

• Feedback during training improves compression quality (strong evidence)
• Skill decay between training events is real and rapid (strong evidence)
• Feedback manikins are effective tools for healthcare training (strong evidence)
• The conceptual case for consumer-grade feedback during between-course practice is sound (reasonable inference)

The category of evidence that would still strengthen the case: outcome research specifically on consumer apps and their impact on bystander CPR quality at real cardiac arrest events. This kind of research is harder to design and execute than controlled-environment training studies, and the field is still building up evidence in this area.

The practical takeaway

If you’ve taken a CPR class, your technique is probably better right now than it’ll be in six months — unless you practice. If you’ve never taken one but watched a video, your technique exists only as conceptual knowledge, not motor skill. In both cases, real-time feedback during practice closes a meaningful gap between what you know and what your hands actually do.

This is the foundation of CPR Coach. Not a replacement for in-person training, but an extension of the well-established feedback principle into the home, where the practice happens between formal courses.